Heterosexual transmission of HIV has been well documented and may represent the major route of virus spread into the general population. The extend of the problem and the factors associated with the heterosexual transmission of HIV are not yet well understood. The purpose of the current proposal is to determine the seroprevalence and seroincidence rates of HIV infection in two separate cohorts, including female prostitutes and sexually active heterosexuals who have sexual contact with prostitutes or who have multiple heterosexual partners. Participants will be recruited from an HIV screening clinic at the medical center. Family Health Centers, Sexually Transmitted Disease Clinics and via advertisement through local newspapers, radio shows, posters and flyers (placers). Evaluation will include an medical history, physical examination, a standardized interview, and laboratory tests. The interview will focus on socioeconomic status, education, occupation, sexual activity and practices, birth control methods, other sexually transmitted diseases, general health, and risk factors for HIV infection. Laboratory tests will include T- cell enumeration, serologic tests for hepatitis, syphilis, cytomegalovirus, chlamydia, and HIV, and cultures for gonorrhea, herpes, and HIV. Participants found to be seronegative for HIV will be followed every 6 months thereafter for 3 years. Seroprevalence and seroincidence rates will be calculated for the entire group and within several subsets. Potential risk factors, modes, and mechanisms for the heterosexual transmission of HIV will be evaluated by comparing individuals who are seropositive for HIV to those who are seronegative. Factors to be evaluated will include length of prostitution, contact with prostitutes, number of sexual partners, number of sexual encounters, types of sexual activity, use of barrier contraceptives (condoms), the presence of other sexually transmitted disease, or potential breaks in epithelial barriers. The presence of HIV p24 antigen, antibody, and presence of virus in vaginal and seminal fluid will be correlated with seroprevalence and seroincidence rates of HIV. Frequency of condom use, proper technique, and leakage or breakage of condoms will also be correlated with seroprevalence and seroincidence rates of HIV. Data analysis will include student's t-test, Fisher's exact test, chi-square test, Wilcoxon rank-sum test, factor analysis of socioeconomic components, and the use of multivariate logistic regression analyses to look for associations, confounding, and interaction.